December 2025 arXiv papers — page 6

Jae Ik Choi, Vahagn Mkhitaryan, Colton Fruhling, Jacob B. Khurgin

We theoretically demonstrate that transparent conducting oxides (TCOs) exhibit oscillatory and sign-reversing dynamic modulation in transmittance on the order of a few optical cycles under extreme electron temperatures, providing a possible explanation for TCO dynamics observed in earlier experiments. We present an inverse-designed multilayer cavity incorpor

Anthony Mikinka

We present Universal Conditional Logic (UCL), a mathematical framework for prompt optimization that transforms prompt engineering from heuristic practice into systematic optimization. Through systematic evaluation (N=305, 11 models, 4 iterations), we demonstrate significant token reduction (29.8%, t(10)=6.36, p < 0.001, Cohen's d = 2.01) with corresponding c

Xiaochi Wu

In this paper, we investigate the existence and characterization of the value for a two-player zero-sum differential game with symmetric incomplete information on a continuum of initial positions and with signal revelation. Before the game starts, the initial position is chosen randomly according to a probability measure with compact support, and neither pla

Siyang Wang, Hanting Li, Wei Li, Jie Hu

Inspired by the remarkable success of autoregressive models in language modeling, this paradigm has been widely adopted in visual generation. However, the sequential token-by-token decoding mechanism inherent in traditional autoregressive models leads to low inference efficiency.In this paper, we propose RadAR, an efficient and parallelizable framework desig

Qingda Hu, Ziheng Qiu, Zijun Xu, Kaizhao Zhang

State ambiguity is common in robotic manipulation. Identical observations may correspond to multiple valid behavior trajectories. The visuomotor policy must correctly extract the appropriate types and levels of information from the history to identify the current task phase. However, naively extending the history window is computationally expensive and may c

Weihang Shen, Yinqiu Chen, Rong Chen, Haibo Chen

The limited HBM capacity has become the primary bottleneck for hosting an increasing number of larger-scale GPU tasks. While demand paging extends capacity via host DRAM, it incurs up to 78x slowdown due to the massive working sets and poor locality of GPU workloads. We observe, however, that GPU memory access patterns are inherently predictable via kernel l

Tanjum Motin Mitul, Md. Masud Mazumder, Md Nahidul Islam Opu, Shaiful Chowdhury

As Software Engineering enters its new era (SE 3.0), AI coding agents increasingly automate software development workflows. However, it remains unclear how exactly these agents recognize and address software energy concerns-an issue growing in importance due to large-scale data centers, energy-hungry language models, and battery-constrained devices. In this

Zhili Huang, Ling Xu, Chao Liu, Weifeng Sun

Automated Program Repair (APR) aims to automatically generate correct patches for buggy programs. Recent approaches leveraging large language models (LLMs) have shown promise but face limitations. Most rely solely on static analysis, ignoring runtime behaviors. Some attempt to incorporate dynamic signals, but these are often restricted to training or fine-tu

Chandler Haight, Svetlana Roudenko, Zhongming Wang

We present a comparative study of classical numerical solvers, such as Petviashvili's method or finite difference with Newton iterations, and neural network-based methods for computing ground states or profiles of solitary-wave solutions to the one-dimensional dispersive PDEs that include the nonlinear Schr\"odinger, the nonlinear Klein-Gordon and the genera

Jan Ambjørn, Yukimura Izawa, Yuki Sato

We study the continuum limit of branched polymers (BPs) with loops coupled to Ising spins at the zero-temperature critical point. It is known that the continuum partition function can be represented by a Hermitian two-matrix model, and we propose a string field theory whose Dyson-Schwinger equation coincides with the loop equation of this continuum matrix mo

Md Nazmus Sakib, Ishika Tarin, Naga Manogna Rayasam, Manas Gaur

In collaborative settings, sustaining momentum and engagement between checkpoints (e.g., meetings) can be challenging, often leading to task drift and reduced preparedness. To address this gap, we developed ReflectEd, an AI-assisted system that supports between-checkpoint reflection through theory-driven prompts with progressively structured levels and mecha

Quanyu Tang, Hao Zhang

Let $\varepsilon>0$ and, for an odd prime $p$, set $$ S_\ell(p):=\sum_{n\le \ell}\left(\frac{n}{p}\right). $$ Define the first-passage time $$ f_\varepsilon(p):=\min\{\ell\ge 1:\ S_\ell(p)<\varepsilon\ell\}. $$ We prove that there exists a constant $c_\varepsilon>0$ such that, as $x\to\infty$, $$ \sum_{p\le x} f_\varepsilon(p)\sim c_\varepsilon \frac{x}{\log

Yanyi Liu, Qingwen Yang, Tiezheng Guo, Feiyu Qu

Hallucinations in Large Language Models (LLMs), defined as the generation of content inconsistent with facts or context, represent a core obstacle to their reliable deployment in critical domains. Current research primarily focuses on binary "detection" approaches that, while capable of identifying hallucinations, fail to provide interpretable and actionable

Md Nahidul Islam Opu, Shahidul Islam, Muhammad Asaduzzaman, Shaiful Chowdhury

LLM-based software engineering is influencing modern software development. In addition to correctness, prior studies have also examined the performance of software artifacts generated by AI agents. However, it is unclear how exactly the agentic AI systems address performance concerns in practice. In this paper, we present an empirical study of performance-re

Mohsen Annabestani, Samira Aghadoost, Anais Rameau, Olivier Elemento

Benign laryngeal voice disorders affect nearly one in five individuals and often manifest as dysphonia, while also serving as non-invasive indicators of broader physiological dysfunction. We introduce a clinically inspired hierarchical machine learning framework for automated classification of eight benign voice disorders alongside healthy controls, using ac

Patrick Iglesias-Zemmour

In Part I, we established the construction of the Prequantum Groupoid for simply connected spaces. This second part extends the theory to arbitrary connected parasymplectic diffeological spaces $(\mathrm{X}, \omega)$. We identify the obstruction to the existence of the Prequantum Groupoid as the non-additivity of the integration of the prequantum form $\math

Dong-Qi Ma, Qing-Xuan Jie, Ya-Dong Hu, Wen-Yi Zhu

Quantum information processing platforms based on array of matter qubits, such as neutral atoms, trapped ions, and quantum dots, face significant challenges in scalable addressing and readout as system sizes increase. Here, we propose the "Volcano" architecture that establishes a new quantum processing unit implementation method based on optical channel mapp

Deepak Babu Piskala

The rapid scaling of deep neural networks and large language models has collapsed the once-clear divide between "research" and "engineering" in AI organizations. Drawing on a qualitative synthesis of public job descriptions, hiring criteria, and organizational narratives from leading AI labs and technology companies, we propose the AI Roles Continuum: a fram

Borui Du, Yumeng Zhang, Christos Masouros, Bruno Clerckx

The orthogonal time frequency space (OTFS) signal is considered a promising solution for high-mobility wireless environments. It manages Doppler effects by utilizing delay-Doppler (DD) domain processing. However, the relatively long OTFS frame duration could introduce considerable sensing or communication latency when radar and communication are performed se

Naoki Ito

This technical report presents a comprehensive study of SDPT3, a widely used open-source MATLAB solver for semidefinite-quadratic-linear programming, which is based on the interior-point method. It includes a self-contained and consistent description of the algorithm, with mathematical notation carefully aligned with the implementation. The aim is to offer a

Qingyu Xu, Runtong Zhang, Zihuan Qiu, Fanman Meng

Object detection in fire rescue scenarios is importance for command and decision-making in firefighting operations. However, existing research still suffers from two main limitations. First, current work predominantly focuses on environments such as mountainous or forest areas, while paying insufficient attention to urban rescue scenes, which are more freque

Lucas A. Souza

We study short-horizon forecasting in financial time series under strict causal constraints, treating the market as a non-stationary stochastic system in which any predictive observable must be computable online from information available up to the decision time. Rather than proposing a machine-learning predictor or a direct price-forecast model, we focus on

Shuyuan Lin, Yu Guo, Xiao Chen, Yanjie Liang

Establishing the correct correspondence of feature points is a fundamental task in computer vision. However, the presence of numerous outliers among the feature points can significantly affect the matching results, reducing the accuracy and robustness of the process. Furthermore, a challenge arises when dealing with a large proportion of outliers: how to ens

Junru Lu, Jiarui Qin, Lingfeng Qiao, Yinghui Li

We introduce Youtu-LLM, a lightweight yet powerful language model that harmonizes high computational efficiency with native agentic intelligence. Unlike typical small models that rely on distillation, Youtu-LLM (1.96B) is pre-trained from scratch to systematically cultivate reasoning and planning capabilities. The key technical advancements are as follows: (

Xingwei Qu, Shaowen Wang, Zihao Huang, Kai Hua

Large Language Models (LLMs) apply uniform computation to all tokens, despite language exhibiting highly non-uniform information density. This token-uniform regime wastes capacity on locally predictable spans while under-allocating computation to semantically critical transitions. We propose $\textbf{Dynamic Large Concept Models (DLCM)}$, a hierarchical lang

Fan Yang

We study the unitary almost Mathieu operator (UAMO), a one-dimensional quasi-periodic unitary operator arising from a two-dimensional discrete-time quantum walk on $\mathbb Z^2$ in a homogeneous magnetic field. In the positive Lyapunov exponent regime $0\le \lambda_1<\lambda_2\le 1$, we establish an arithmetic localization statement governed by the frequency

Yuchen Shi, Yuzheng Cai, Siqi Cai, Zihan Xu

Existing Large Language Model (LLM) agent frameworks face two significant challenges: high configuration costs and static capabilities. Building a high-quality agent often requires extensive manual effort in tool integration and prompt engineering, while deployed agents struggle to adapt to dynamic environments without expensive fine-tuning. To address these

Yuya Miyaoka, Masaki Inoue, Kengo Urata, Shigeaki Harada

This paper proposes a chat-driven network management framework that integrates natural language processing (NLP) with optimization-based virtual network allocation, enabling intuitive and reliable reconfiguration of virtual network services. Conventional intent-based networking (IBN) methods depend on statistical language models to interpret user intent but

Zheyu Shi, Dong Qiu, Shanlong Yu

This paper proposes a group deliberation oriented multi-agent conversational model to address the limitations of single large language models in complex reasoning tasks. The model adopts a three-level role division architecture consisting of generation, verification, and integration. An opinion generation agent produces diverse reasoning perspectives, an evi

Qing-He Ni, Rujia Wang, Tianyang Xie, Jingxin Zhang

We present the ExoAtom database, www.exomol.com/exoatom, an extension of the ExoMol database to provide atomic line lists in the ExoMol format. ExoAtom is designed for detailed astrophysical, planetary, and laboratory applications. ExoAtom currently includes atomic data for 80 neutral atoms and 74 singly charged ions. These data are extracted from both the N

Yiwen Shao, Wei Liu, Jiahong Li, Tianzi Wang

Extending large language models (LLMs) to the speech domain has recently gained significant attention. A typical approach connects a pretrained LLM with an audio encoder through a projection module and trains the resulting model on large-scale, task-specific instruction-tuning datasets. However, curating such instruction-tuning data for specific requirements

Kwangok Seo, Johan Lim, Kaiwen Wang, Dohwan Park

In this paper, we address the normal mean inference problem, which involves testing multiple means of normal random variables with heteroscedastic variances. Most existing empirical Bayes methods for this setting are developed under restrictive assumptions, such as the scaled inverse-chi-squared prior for variances and unimodality for the non-null mean distr

Zhiyuan Geng

This paper studies minimizing solutions to a two dimensional Allen-Cahn system on the upper half plane, subject to Dirichlet boundary conditions, \begin{equation*} \Delta u-\nabla_u W(u)=0, \quad u: \mathbb{R}_+^2\to \mathbb{R}^2,\ u=u_0 \text{ on } \partial \mathbb{R}_+^2, \end{equation*} where $W: \mathbb{R}^2\to [0,\infty)$ is a multi-well potential. We g

Dong Qiu, Duo Xu, Limengxi Yue

Large Language Models (LLMs) perform well in language tasks but often lack collaborative awareness and struggle to optimize global performance in multi-agent settings. We present a reinforcement learning-augmented LLM agent framework that formulates cooperation as a decentralized partially observable Markov decision process (Dec-POMDP) and adopts centralized

Cesar A. Ipanaque Zapata, Martha O. Gonzales Bohorquez

We present the notion of injective hom-complexity, leading to a connection between the covering number of a group and the sectional number of a group homomorphism, and provide estimates for computing this invariant.

Yusuke Nemoto, Takuya Yamauchi

In this paper, we construct certain rational or integral elements in the motivic cohomology of superelliptic curves which are quotient curves of abelian coverings of $\mathbb{P}^1$ minus $n+2$ points, and prove that these elements are non-trivial by expressing their regulators in terms of Appell-Lauricella hypergeometric functions. We also check that such el

Yujun Ju

Motivated by the notion of intermediate dimensions introduced by Falconer et al., we introduce a continuum of topological entropies that are intermediate between the (Bowen) topological entropy and the lower and upper capacity topological entropies. This is achieved by restricting the families of allowable covers in the definition of topological entropy by r

Panquan Yang, Junfei Huang, Zongzhangbao Yin, Yingsong Hu

3D visual grounding aims to localize the object in 3D point cloud scenes that semantically corresponds to given natural language sentences. It is very critical for roadside infrastructure system to interpret natural languages and localize relevant target objects in complex traffic environments. However, most existing datasets and approaches for 3D visual gro

Ryo Ohashi, Hiroyasu Abe, Fumitake Sakaori

Non-negative matrix factorization (NMF) is widely used as a feature extraction technique for matrices with non-negative entries, such as image data, purchase histories, and other types of count data. In NMF, a non-negative matrix is decomposed into the product of two non-negative matrices, and the approximation accuracy is evaluated by a loss function. If th

Zheng Liu, Jinchao Zhu, Gao Huang

Low-rank adaptation (LoRA) has achieved remarkable success in fine-tuning pre-trained vision transformers for various downstream tasks. Existing studies mainly focus on exploring more parameter-efficient strategies or more effective representation learning schemes. However, these methods either sacrifice fine-tuning performance or introduce excessive trainab

Weixuan Chen, Qianqian Yang, Yuanyuan Jia, Junyu Pan

Semantic communication (SemCom) has emerged as a promising paradigm for future wireless networks by prioritizing task-relevant meaning over raw data delivery, thereby reducing communication overhead and improving efficiency. However, shifting from bit-accurate transmission to task-oriented delivery introduces new security and privacy risks. These include sem

Alex L. Zhang, Tim Kraska, Omar Khattab

We study allowing large language models (LLMs) to process arbitrarily long prompts through the lens of inference-time scaling. We propose Recursive Language Models (RLMs), a general inference paradigm that treats long prompts as part of an external environment and allows the LLM to programmatically examine, decompose, and recursively call itself over snippet

Hung D. Nguyen, Kihoon Seong

We study the long-time mixing behavior of the stochastic nonlinear Schr\"odinger equation in $\mathbb{R}^d$, $d\le 3$. It is well known that, under a sufficiently strong damping force, the system admits unique ergodicity, although the rate of convergence toward equilibrium has remained unknown. In this work, we address the mixing property in the regime of la

Slim Ibrahim, Tatsuya Miura, Carlos Román, Ikkei Shimizu

We study a variational problem for the Landau--Lifshitz energy with Dzyaloshinskii--Moriya interactions arising in 2D micromagnetics, focusing on the Bogomol'nyi regime. We first determine the minimal energy for arbitrary topological degree, thereby revealing two types of phase transitions consistent with physical observations. In addition, we prove the uniq

Manish Bhatt, Adrian Wood, Idan Habler, Ammar Al-Kahfah

Production LLM agents with tool-using capabilities require security testing despite their safety training. We adapt Go-Explore to evaluate GPT-4o-mini across 28 experimental runs spanning six research questions. We find that random-seed variance dominates algorithmic parameters, yielding an 8x spread in outcomes; single-seed comparisons are unreliable, while

Hiroki Kajiura, Makoto Matsumoto

A difference set with parameters $(v, k, \lambda)$ is a subset $D$ of cardinality $k$ in a finite group $G$ of order $v$, such that the number $\lambda$ of occurrences of $g \in G$ as the ratio $d^{-1}d'$ in distinct pairs $(d, d')\in D\times D$ is independent of $g$. We prove the nonexistence of $(120, 35, 10)$-difference sets, which has been an open proble

Wenxuan Xie, John C. Schotland

We present a framework to investigate the collective properties of atomic lattices in one, two, and three dimensions. We analyze the single-photon band structure and associated atomic decay rates, revealing a fundamental dependence on dimensionality. One- and two-dimensional arrays are shown to be inherently radiative, exhibiting band gaps and decay rates th

Jozsef Solymosi, Josh Zahl

John E. Littlewood posted the question {\em ``Is it possible in 3-space for seven infinite circular cylinders of unit radius each to touch all the others? Seven is the number suggested by counting constants.''} Boz\'oki, Lee, and R\'onyai constructed a configuration of 7 mutually touching unit cylinders. The best-known upper bounds show that at most 10 unit

Zhongyi Wang, Tengjie Lin, Mingshuai Chen, Haokun Li

Fully automated verification of large-scale software and hardware systems is arguably the holy grail of formal methods. Large language models (LLMs) have recently demonstrated their potential for enhancing the degree of automation in formal verification by, e.g., generating formal specifications as essential to deductive verification, yet exhibit poor scalab

Nhut Le, Maryam Rahnemoonfar

The increasing frequency of natural disasters poses severe threats to human lives and leads to substantial economic losses. While 3D semantic segmentation is crucial for post-disaster assessment, existing deep learning models lack datasets specifically designed for post-disaster environments. To address this gap, we constructed a specialized 3D dataset using

Fuyu Dong, Ke Li, Di Wang, Nan Luo

Change detection visual question answering (CDVQA) requires answering text queries by reasoning about semantic changes in bi-temporal remote sensing images. A straightforward approach is to boost CDVQA performance with generic vision-language models via supervised fine-tuning (SFT). Despite recent progress, we observe that a significant portion of failures d

Grant Davis, James K. Freericks

Using a global rotation by theta about the z-axis in the spin sector of the Jordan-Wigner transformation rotates Pauli matrices X and Y in the x-y-plane, while it adds a global complex phase to fermionic quantum states that have a fixed number of particles. With the right choice of angles, this relates expectation values of Pauli strings containing products

Sunay Joshi, Yan Sun, Hamed Hassani, Edgar Dobriban

As generative AI systems are increasingly deployed in real-world applications, regulating multiple dimensions of model behavior has become essential. We focus on test-time filtering: a lightweight mechanism for behavior control that compares performance scores to estimated thresholds, and modifies outputs when these bounds are violated. We formalize the prob

Xian-Peng Zhang, Xiangrong Wang, Yugui Yao

Magnetoresistance (MR) in magnetic materials arises from spin-exchange coupling between local moments and itinerant electrons, representing a challenging many-body open-quantum problem. Here we develop a comprehensive microscopic theory of MR within an open-quantum-system framework by solving the Liouville-von Neumann equation for a hybrid system of free ele

Yongwei Li, Rustem Khasanov, Stephen P. Cottrell, Naoki Kikugawa

The nature of the broken time reversal symmetry (BTRS) state in Sr$_2$RuO$_4$ remains elusive, and its relation to superconductivity remains controversial. There are various universal predictions for the BTRS state when it is associated with a multicomponent superconducting order parameter. In particular, in the BTRS superconducting state, spontaneous fields

Jun-Ru Chen, Yong-Qiang Wang

In this work, we investigate the five-dimensional Proca star under gravity with the infinite tower of higher curvature corrections. We find that when the coupling constant exceeds a critical value, solutions with a frequency approaching zero appear. In the finite-order corrections case $n=2$ (Gauss-Bonnet gravity), the matter field and energy density diverge

Rahul Gulia, Ashish Sheikh, Feyisayo Favour Popoola, Serisha Vadlamudi

This paper addresses a fundamental physical layer conflict in hybrid Wireless Sensor Networks (WSNs) between high-throughput primary communication and the stringent power envelope requirements of passive backscatter sensors. We propose a Backscatter-Constrained Transmit Antenna Selection (BC-TAS) framework, a per-subcarrier selection strategy for multi-anten

Kenichi Ito, Tomoya Tagawa

We discuss propagation of space-time singularities for the quantum harmonic oscillator with time-dependent metric and potential perturbations. Reformulating the quasi-homogeneous wave front set according to Lascar (1977) in a semiclassical manner, we obtain a characterization of its appearance in comparison with the unperturbed system. The idea of our proof

Igor N. Tunev

Let the root of the word $w$ be the smallest prefix $v$ of $w$ such that $w$ is a prefix of $vvv...$. $per(w)$ is the length of the root of $w$. For any $n\ge5$, an $n$-ary threshold word is a word $w$ such that for any factor (subword) $v$ of $w$ the condition $\frac{|v|}{per(v)}\le\frac{n}{n-1}$ holds. Dejean conjecture (completely proven in 2009) states f

Shanyu Han, Yangbo He, Yang Liu

We propose a novel framework for risk-sensitive reinforcement learning (RSRL) that incorporates robustness against transition uncertainty. We define two distinct yet coupled risk measures: an inner risk measure addressing state and cost randomness and an outer risk measure capturing transition dynamics uncertainty. Our framework unifies and generalizes most

Hao Li, Tanausú del Pino Alemán, Javier Trujillo Bueno

The polarization of the Mg II h and k lines holds significant diagnostic potential for measuring chromospheric magnetic fields, which are crucial for understanding the physical processes governing the energy transport and dissipation in the solar upper atmosphere, as well as the subsequent heating of the chromosphere and corona. The Chromospheric Layer Spect

Rui Mao, Pei Yuan, Jonathan Allcock, Shengyu Zhang

The QAOA has been the subject of intense study over recent years, yet the corresponding Dynamical Lie Algebra (DLA)--a key indicator of the expressivity and trainability of VQAs--remains poorly understood beyond highly symmetric instances. An exponentially scaling DLA dimension is associated with the presence of so-called barren plateaus (BP) in the optimiza

Zahid Ullah, Jihie Kim

Knee osteoarthritis (KOA) is a leading cause of disability worldwide, and accurate severity assessment using the Kellgren Lawrence (KL) grading system is critical for clinical decision making. However, radiographic distinctions between early disease stages, particularly KL1 and KL2, are subtle and frequently lead to inter-observer variability among radiologi

Xuanang Hu, Vladimir V. Ulyanov, Hanchao Wang

We give a dual, separated-tree formulation of Latala's majorizing measure theorem for canonical processes with log-concave tails. Under the same assumptions as in Latala's characterization, we introduce parameterized separation trees and prove that the expected supremum is equivalent, up to universal constants, to the corresponding tree functional. We also d

Javad Mashreghi, Mostafa Nasri, Prateek Kumar Vishwakarma

Convolution admits a natural formulation as a functional operation on matrices. Motivated by the functional and entrywise calculi, this leads to a framework in which convolution defines a matrix transform that preserves positivity. Within this setting, we establish results parallel to the classical theories of P\'olya--Szeg\H{o}, Schoenberg, Rudin, Loewner,

Zhenyu Zhang, Xiaoxia Wu, Zhongzhu Zhou, Qingyang Wu

Large Language Models (LLMs) often rely on long chain-of-thought (CoT) reasoning to solve complex tasks. While effective, these trajectories are frequently inefficient, leading to high latency from excessive token generation, or unstable reasoning that alternates between underthinking (shallow, inconsistent steps) and overthinking (repetitive, verbose reason

Lei Li, Yanqing Xu, Tenghao Cai, Tsung-Hui Chang

With great flexibility to adjust antenna positions, pinching antennas (PAs) are promising for alleviating large-scale attenuation in wireless networks. In this work, we investigate the antenna positioning and beamforming (AP-BF) design in a multi-PA multi-user system under probabilistic light-of-sight (LoS) blockage and formulate a power minimization problem

Zhihao Wang, Yiyang Liu, Ting Wang, Zhiyuan Liu

Large language models (LLMs) are increasingly used in medicine and clinical workflows, yet we know little about their decision and affective profiles. Taking a historically informed outlook on the future, we treated successive OpenAI models as an evolving lineage and compared them with humans in a gambling task with repeated happiness ratings. Computational

Hongseok Oh, Wonseok Hwang, Kyoung-Woon On

We introduce the Korean Canonical Legal Benchmark (KCL), a benchmark designed to assess language models' legal reasoning capabilities independently of domain-specific knowledge. KCL provides question-level supporting precedents, enabling a more faithful disentanglement of reasoning ability from parameterized knowledge. KCL consists of two components: (1) KCL

Md Hasan Saju, Austin Page, Akramul Azim, Jeff Gardiner

Security Information and Event Management (SIEM) systems are essential for large enterprises to monitor their IT infrastructure by ingesting and analyzing millions of logs and events daily. Security Operations Center (SOC) analysts are tasked with monitoring and analyzing this vast data to identify potential threats and take preventive actions to protect ent

Shiqi Kuang, Zhao Tian, Tao Xiao, Dong Wang

Large language models (LLMs) have achieved remarkable progress in code generation, largely driven by the availability of high-quality code datasets for effective training. To further improve data quality, numerous training data optimization techniques have been proposed; however, their overall effectiveness has not been systematically evaluated. To bridge th

Elvis Cabrera, Jyrko Correa

Lattices induced by coverings arise naturally in matroid theory and combinatorial optimization, providing a structured framework for analyzing relationships between independent sets and closures. In this paper, we explore the structural properties of such lattices, with a particular focus on their rank structure, covering relations, and enumeration of elemen

Alexey Cheskidov, Qirui Peng

We construct solutions to the three-dimensional Euler equations exhibiting anomalous dissipation in finite time through a vanishing viscosity limit. Inspired by \cite{BDL23} and \cite{cheskidov2023dissipation}, we extend the \(2\frac{1}{2}\)-dimensional constructions and establish an Onsager-critical energy criterion adapted to such flows, showing its sharpn

Volkan Yildiz

We introduce Heraclitean Dialectical Concept Spaces (HDCS), a topological framework for modelling how concepts evolve. Concepts are represented as open regions generated by neighbourhoods in a feasible family, and their relationships are organised through overlaps and channel ideals. New concepts emerge from remainders where existing structures fail to fit t

Siqi Li, Xinyu Cai, Jianbiao Mei, Nianchen Deng

Recent multimodal large language models (MLLMs) show strong capabilities in visual-language reasoning, yet their performance on ultra-high-resolution imagery remains largely unexplored. Existing visual question answering (VQA) benchmarks typically rely on medium-resolution data, offering limited visual complexity. To bridge this gap, we introduce Ultra-high-

Hannes Vandecasteele, Nicholas Karris, Alexander Cloninger, Ioannis G. Kevrekidis

Timesteppers constitute a powerful tool in modern computational science and engineering. Although they are typically used to advance the system forward in time, they can also be viewed as nonlinear mappings that implicitly encode steady states and stability information. In this work, we present an extension of the matrix-free framework for calculating, via t

Joonggwon Kim, Dogyun Byeon

The release of hydrophilic solutes from micron scale particulate formulations can be understood as an interfacial transport problem in which diffusion across a heterogeneous amphiphilic coating competes with dissolution and convective removal in the surrounding medium. Here we reinterpret a glycerin fatty acid ester (GFAE) coated thiamine (vitamin B1) micro

Wenrui Liu, Zixiang Liu, Elsie Dai, Wenhan Yu

Large Language Models (LLMs) are increasingly serving as autonomous agents, and their utilization of external tools via the Model Context Protocol (MCP) is considered a future trend. Current MCP evaluation sets suffer from issues such as reliance on external MCP services and a lack of difficulty awareness. To address these limitations, we propose MCPAgentBen

Shunbo Jia, Caizhi Liao

Deep learning models for Electrocardiogram (ECG) diagnosis have achieved remarkable accuracy but exhibit fragility against adversarial perturbations, particularly Smooth Adversarial Perturbations (SAP) that mimic biological morphology. Existing defenses face a critical dilemma: Adversarial Training (AT) provides robustness but incurs a prohibitive computatio

Xue-Zhao Chang, HouJun Lü, Jia-Ming Chen, En-Wei Liang

GRB 161117A is a long-duration GRB with three main overlapping peaks. By analyzing the time-resolved spectra of its data observed with the Gamma-Ray Burst Monitor (GBM) on board the Fermi mission, we find that the spectral evolution shows a transition from thermal (single BB) to hybrid (PL$+$BB), and finally to non-thermal (Band and CPL) emissions. Such a tr

Prajwal Chinchmalatpure, Suyash Chinchmalatpure, Siddharth Chavan

Generative audio technologies now enable highly realistic voice cloning and real-time voice conversion, increasing the risk of impersonation, fraud, and misinformation in communication channels such as phone and video calls. This study investigates real-time detection of AI-generated speech produced using Retrieval-based Voice Conversion (RVC), evaluated on

David Gros, Prem Devanbu

Large Language models (LLMs) can generate complicated source code from natural language prompts. However, LLMs can generate output that deviates from what the user wants, requiring supervision and editing. To support this process, we offer techniques to localize where generations might be misaligned from user intent. We first create a dataset of "Minimal Int

Japan K. Patel, Barry D. Ganapol, Anthony Magliari, Matthew C. Schmidt

We present a genetic programming approach to automatically discover convergence acceleration methods for discrete ordinates solutions of neutron transport problems in slab geometry. Classical acceleration methods such as Aitken's delta-squared and Wynn epsilon assume specific convergence patterns and do not generalize well to the broad set of transport probl

Shuvro Chowdhury, Jasper Pieterse, Navid Anjum Aadit, Shaila Niazi

Neural quantum states efficiently represent many-body wavefunctions with neural networks, but the cost of Monte Carlo sampling limits their scaling to large system sizes. Here we address this challenge by combining sparse Boltzmann machine architectures with probabilistic computing hardware. We implement a probabilistic computer on field-programmable gate ar

Muhammad Abdullahi Said, Muhammad Sammani Sani

As Large Language Models (LLMs) integrate into critical global infrastructure, the assumption that safety alignment transfers zero-shot from English to other languages remains a dangerous blind spot. This study presents a systematic audit of three state of the art models (GPT-5.1, Gemini 3 Pro, and Claude 4.5 Opus) using HausaSafety, a novel adversarial data

Xueyan Li, Yingyi Xue, Mengjie Jiang, Qingzi Zhu

Generating humorous memes is a challenging multimodal task that moves beyond direct image-to-caption supervision. It requires a nuanced reasoning over visual content, contextual cues, and subjective humor. To bridge this gap between visual perception and humorous punchline creation, we propose HUMOR}, a novel framework that guides VLMs through hierarchical r

G. Panotopoulos, Á. Rincón, I. Lopes

In this study we obtain interior solutions and investigate structural properties of isotropic compact stars in the framework of four-dimensional regularized Einstein-Gauss-Bonnet (4DEGB) gravity. For stellar matter content, we adopt a widely used quark-matter model that approximates a realistic equation of state (EoS). By numerically integrating the modified

Bohm-Jung Yang

For decades, ``geometry" in band theory has largely meant Berry phase and Berry curvature-quantities that reshape semiclassical dynamics and underpin modern topological matter. Yet the full geometric content of a Bloch band is richer and encoded in the quantum geometric tensor (QGT), whose imaginary part is the Berry curvature and whose real part is the quan

Jindi Zhong, Congyaohui Yin, Zhaorong Zhang, Huanshui Zhang

This paper proposes a novel second-order optimization algorithm based on the Optimal Control Principle (OCP), applicable to large-scale optimization problems in neural network training. The algorithm has a computational complexity of O(d) and strong robustness. Extensive experiments on multiple benchmarks demonstrate the significant superiority of the propos

Tomoya Yamanokuchi, Alberto Bacchin, Emilio Olivastri, Ryotaro Arifuku

In this work, we address the limitation of surface fitting-based grasp planning algorithm, which primarily focuses on geometric alignment between the gripper and object surface while overlooking the stability of contact point distribution, often resulting in unstable grasps due to inadequate contact configurations. To overcome this limitation, we propose a n

Maor Ben-Shahar, Lucile Cangemi, Henrik Johansson

We develop a quadratic-in-Riemann worldline action for a Kerr black hole at infinite spin orders by matching to a proposed tree-level Kerr Compton amplitude, originally obtained from higher-spin QFT considerations. A worldline action is an effective theory, and as such the tree-level matching needs to be corrected by loop effects, including UV counter terms,

Samuel Dentan

We establish a partial extension of the Poincar\'e duality theorem of Jell-Rau-Shaw to tropical hypersurfaces arising from non-primitive triangulations. We introduce a notion of level of primitivity for triangulations of lattice polytopes and show that tropical hypersurfaces satisfy a partial form of Poincar\'e duality determined by this level. This notion o

Manikanta Kotthapalli, Banafsheh Rekabdar

The exponential growth of video traffic has placed increasing demands on bandwidth and storage infrastructure, particularly for content delivery networks (CDNs) and edge devices. While traditional video codecs like H.264 and HEVC achieve high compression ratios, they are designed primarily for pixel-domain reconstruction and lack native support for machine l

Yuto Nogata

For a finite group $G$, let $a_n(G)$ be the number of subgroups of order $n$ and define $\zeta_G(s)=\sum_{n\ge 1} a_n(G)n^{-s}$. Examples are known of non-isomorphic finite groups with the same group zeta function. However, no general criterion is known for when two finite groups have the same group zeta function. Fix integers $m,n\ge 1$ and a prime $p$, and

Yuma Ichikawa, Yoshihiko Fujisawa, Yudai Fujimoto, Akira Sakai

For extreme low-bit quantization of large language models (LLMs), Double Binary Factorization (DBF) is attractive as it enables efficient inference without sacrificing accuracy. However, the scaling parameters of DBF are too restrictive; after factoring out signs, all rank components share the same magnitude profile, resulting in performance saturation. We p

M. J. Gerard, M. J. Pueschel, S. Stewart, H. O. M. Hillebrecht

There is currently no agreed-upon methodology for characterizing a stellarator magnetic field geometry, and yet modern stellarator designs routinely attain high levels of magnetic-field quasi-symmetry through careful flux-surface shaping. Here, we introduce a general method for computing the shape of an ideal-MHD equilibrium that can be used in both axisymme

Bo Li, Zijun Chen, Haiwang Zhong, Di Cao

In wide-area measurement systems (WAMS), phasor measurement unit (PMU) measurement is prone to data missingness due to hardware failures, communication delays, and cyber-attacks. Existing data-driven methods are limited by inadaptability to concept drift in power systems, poor robustness under high missing rates, and reliance on the unrealistic assumption of

Tasman Fell

We construct a diagrammatic categorification of the spherical module over the Hecke algebra. We establish a basis for the morphism spaces of this category, and prove that it is equivalent to an existing algebraic spherical category.

Samuel. H. DAmbrosia, Adrianne Zhong, Michael R. DeWeese

Linear response theory has found many applications in statistical physics. One of these is to compute minimal-work protocols that drive nonequilibrium systems between different thermodynamic states, which are useful for designing engineered nanoscale systems and understanding biomolecular machines. We compare and explore the relationships between linear-resp

Cyril Metzger, Alec L. Emser, Brendon C. Rose, Konrad W. Lehnert

We present experimental evidence of a thermally-driven amplitude-frequency nonlinearity in a thin-film quartz phononic crystal resonator at millikelvin temperatures. The nonlinear response arises from the coupling of the mechanical mode to an ensemble of microscopic two-level system defects driven out of equilibrium by a microwave drive. In contrast to the c